Posts Tagged ‘Scalability’
[reClojure2025] Datomic at Nubank: Simplicity Scaled
Lecturer
Jordan Miller is a Staff Software Engineer at Nubank, where she is a core member of the Datomic team, specifically focusing on developer success and the evolution of the database product. Jordan is a prominent figure in the Clojure community, serving as the co-founder of Clojure Camp, a co-author of Clojure Brain Teasers, and the producer/host of the Lost In Lambduhhs podcast. Her professional background is marked by a deep passion for functional programming and a commitment to making complex technical systems accessible to a wider audience.
Abstract
Since its acquisition by Nubank in 2020, Datomic has transitioned from a specialized tool for discerning architects to the foundational storage layer of one of the world’s largest independent banks. This article explores the symbiotic relationship between Nubank’s mission to simplify banking and Datomic’s philosophy of data-driven simplicity. We analyze the technical challenges of scaling Datomic to serve over 114 million customers, focusing on innovations such as “implicit partitions” and the architectural shifts required to maintain a global-scale financial system. Furthermore, we examine how the “Datomic-as-a-Service” model within Nubank empowers thousands of developers to build resilient, high-performance microservices while maintaining the lowest cost-to-serve in the banking industry.
Simplicity as a Mission: The Nubank and Datomic Alignment
The acquisition of Cognitect (the creators of Clojure and Datomic) by Nubank was more than a corporate transaction; it was a meeting of shared values. Nubank was founded with the explicit goal of reducing the extreme complexity and high costs of the Brazilian banking sector. At the time, millions of Brazilians were unbanked, and interest rates for personal loans often exceeded 100% per year. Nubank sought to disrupt this by building a bank that was digital-first, customer-centric, and architecturally simple.
Datomic, with its focus on immutability, “database-as-a-value,” and the separation of reads from writes, provided the perfect technical backbone for this mission. Jordan highlights that the alignment on “simplicity” allowed the partnership to flourish. Today, Nubank serves 114.3 million users with an 83% activity rate, largely fueled by word-of-mouth referrals. The bank’s ability to scale rapidly while maintaining the lowest cost-to-serve globally is a direct consequence of an architecture that prioritizes developer productivity and system reliability through Datomic.
Architectural Innovation: Scaling the Immutable Store
Scaling a database to handle the needs of over 100 million customers requires rethinking traditional storage patterns. Within Nubank, Datomic is not just a database; it is a managed service provided to over 3,000 developers working across 1,000 microservices. This “Datomic-as-a-Service” (DaaS) model ensures that individual product teams do not need to worry about provisioning or maintaining their own database clusters.
A key innovation discussed by Jordan is the concept of implicit partitions. In standard Datomic, partitions are a way to group related data for better locality and performance. However, as the number of services and the volume of data grew at Nubank, manual partition management became a bottleneck. The Datomic team developed implicit partitions to automate this process, allowing the system to handle massive growth without manual intervention. This technical advancement is part of a broader effort to make Datomic “invisible” to the developer, allowing them to focus on business logic rather than storage mechanics.
The Power of “Database-as-a-Value”
The fundamental innovation of Datomic—treating the database as an immutable value—is what enables Nubank’s high-scale operations. In a traditional SQL database, “the” database is a moving target that changes with every update. In Datomic, a process can take a “snapshot” of the database at a specific point in time and perform complex queries against that static value. This eliminates many of the concurrency issues found in traditional systems and makes it trivial to implement features like “time-travel” auditing, which is critical for financial compliance.
Developer Success and the Future of Datomic
Jordan’s role at Nubank emphasizes that the success of a technology is as much about the people using it as it is about the code itself. The Datomic team at Nubank is committed to “Developer Success,” which involves building tools, documentation, and a culture that helps engineers leverage the database effectively. This includes the creation of a “Datomic Workshop” and active engagement with the Clojure community to ensure that Datomic continues to evolve as a world-class product.
The results of this focus are evident in Nubank’s operational efficiency. By providing a unified, simple, and powerful storage abstraction, Nubank has managed to grow by an average of 13 million customers per year since 2015 while keeping its infrastructure costs significantly lower than traditional banks. This demonstrates that functional programming and immutable databases are not just academic curiosities but are viable, high-performance solutions for the world’s most demanding financial environments.
Links:
[GopherConUK2025] How Just Eat Uses Tooling to Deploy Go Micro-services in Minutes
Lecturer
Ainsley Clark is a Senior Software Engineer at Just Eat, working within the Jet Connect team. He has been with the organisation for just over two years and maintains a strong professional focus on Go. His work centres on the design and operation of the internal microservice development toolkit that underpins large-scale order and menu processing across hundreds of partner integrations.
Abstract
This article describes the evolution and capabilities of GoKit, the internal microservice development toolkit developed by Just Eat’s Jet Connect team. Confronted with the maintenance burden of hundreds of independently forked services, the team replaced a template-based approach with a centralised code-generation and infrastructure-as-code system. The resulting tool scaffolds services, generates event consumers and producers, provisions cloud resources, produces continuous-integration workflows and emits operational metrics with minimal engineer effort. A concrete pizza-order example illustrates how a fully instrumented, event-driven service can be created and extended in minutes rather than days, allowing engineers to concentrate on business logic rather than infrastructure boilerplate.
From Monolith and Templates to a Centralised Toolkit
Jet Connect began life as Flight, an integration platform founded in 2013. Its purpose is to unify order and menu processing across restaurants, groceries and electronics partners so that a single point-of-sale interaction replaces multiple device-specific payloads. Today the platform serves approximately 731 000 partners across seventeen countries and processes hundreds of millions of orders annually. The Jet Connect engineering group itself comprises roughly fifty people and owns more than one hundred Go microservices.
The journey to that estate began with a PHP monolith that became increasingly difficult to change. Integrations were subsequently extracted into TypeScript services that spoke gRPC to the monolith. A GitHub template repository accelerated the creation of new integrations: an engineer could fork the template and obtain environment files, Helm charts and build workflows in minutes. The approach delivered speed in the early days yet exacted a heavy maintenance cost. A bug fix or standardisation change had to be applied manually to every fork. Inconsistent folder structures and coding patterns made context-switching expensive. End-to-end testing was absent, reducing confidence in releases.
The decision was therefore taken to move the entire integration layer to Go and to replace the template model with a purpose-built toolkit. The requirements were exacting: generate or update a repository in seconds; treat OpenAPI documentation as a first-class artefact; allow an engineer to subscribe to events by writing only a handler; hide infrastructure details such as databases, event buses and object storage; auto-generate continuous-integration and continuous-delivery pipelines; guarantee safe production deployment on merge to main; support capability testing across the full service graph; and provide a consistent local development experience with minimal setup. The resulting system is known internally as GoKit.
Scaffolding, Event Handling and Infrastructure as Code
GoKit meets those requirements through a combination of code generation and a declarative service description. The command gokit new <service-name> presents a short interactive questionnaire: does the service consume events, produce events, expose an HTTP server, require a database or object store? On the basis of the answers it scaffolds a consistent folder structure whose most important directories are cmd/app (containing a generated main) and internal (the sole location for custom business logic). A file named service.json acts as the single source of truth for infrastructure; it is consumed by Terraform templates that provision the necessary cloud resources.
A typical event handler receives an HTTP client, an event-bus producer and the incoming event (typed via reflection performed by GoKit). After performing domain work—calling a partner API, writing analytics, storing a payload—the handler emits a successor event. Registration consists of a single method call that associates the handler with a concrete event type. A subsequent gokit update rewrites service.json, updates generated code and refreshes documentation. The resulting README lists every consumed and produced topic, giving any engineer an immediate, accurate overview of the service’s responsibilities without requiring manual documentation effort.
Because the same service.json also drives resource provisioning, adding a DynamoDB table or an S3 bucket is a matter of inserting a few lines of JSON and re-running the update command. Generated interfaces for read/write operations and object-store upload/download encourage dependency injection and make unit testing straightforward. Should the underlying technology later change (for example from DynamoDB to PostgreSQL), only the implementation behind the interface needs to be altered; every consuming service continues to compile and run unchanged. The same mechanism supports vertical and horizontal scaling declarations—minimum replica counts, CPU and memory class—again expressed as simple keys in service.json.
Continuous Integration, Capability Testing and Observability
All continuous-integration workflows are themselves generated by GoKit. A change to a workflow template is made once in the central repository; every service inherits the update the next time gokit update is run. Capability tests spin up the entire relevant service graph under Docker, inject a correlation identifier into every request and event, and assert final outcomes. The approach scales to capability chains that involve ten or twenty intermediate services and provides high confidence before production deployment. Drift detection prevents accidental divergence: if an engineer edits a generated file without running the update command, continuous integration fails the pull request. Version pinning of GoKit itself ensures that services cannot merge while they lag behind a required toolkit version, producing a natural, incremental migration path across the estate.
Operational metrics appear automatically. Grafana dashboards display event rates, HTTP status codes, latency distributions, DynamoDB read/write performance and S3 operation counts without any additional instrumentation code. The same consistency that simplifies development also simplifies observation. When a product request arrives for analytics storage or an order-ready notification endpoint, the engineer adds a handful of lines to service.json or an OpenAPI specification, runs the update command, implements a short handler, and obtains a fully instrumented, tested and documented service ready for review.
Outcomes and Lessons
In the six years of its existence GoKit has allowed the Jet Connect team to update the entire service estate with a single pull request, to maintain uniform folder structures that eliminate costly context switches, and to keep documentation accurate without manual effort. Engineers spend their time on business problems—partner-specific payloads, analytics requirements, notification flows—rather than on Terraform, Helm or continuous-integration boilerplate. The toolkit has been instrumental in scaling the platform to hundreds of millions of orders while keeping the cognitive load on individual developers manageable.
The most transferable lessons are structural rather than technological. A single source of truth for service shape, aggressive code generation of repetitive artefacts, enforcement of consistency through continuous integration, and the deliberate abstraction of infrastructure behind stable interfaces together convert the chronic maintenance burden of a large microservice estate into a manageable, largely automated background process. The result is that complex event-driven workflows can be designed, implemented, tested and deployed in minutes rather than days, freeing engineering capacity for the differentiated work that actually delivers value to partners and customers.
The pizza-service walkthrough, though simplified for presentation, captures the essential rhythm of day-to-day work. An engineer begins with a short questionnaire, receives a fully structured repository, writes a handful of domain-specific lines, runs an update command, and obtains a service that already possesses continuous-integration pipelines, infrastructure declarations, metrics dashboards and capability-test scaffolding. Subsequent product requests—analytics storage, partner callbacks, new event types—are accommodated by small, localised edits rather than by the recreation of entire deployment artefacts. The cognitive load remains focused on the business problem; the mechanical work of packaging, provisioning and observing has been systematically removed from the critical path.
The same pattern scales. Whether the service handles a single partner’s CSV feed or participates in a multi-stage order-reconciliation flow involving a dozen intermediate services, the surrounding machinery remains identical. Consistency of structure, generation of repetitive artefacts and enforcement of hygiene through continuous integration together produce an estate that can grow without a proportional increase in operational friction. That is the practical payoff of the investment in a centralised toolkit.
Looking ahead, the same principles that make GoKit effective inside Just Eat are portable to any organisation that maintains a large collection of similarly structured services. The precise implementation will differ—different cloud providers, different event buses, different continuous-integration systems—but the underlying ideas remain constant: centralise the definition of service shape, generate everything that can be generated, enforce consistency automatically, and keep the engineer’s attention on the business problem. When those ideas are applied with discipline, the cost of creating and operating microservices falls dramatically, and the organisation’s capacity to respond to new partner or product requirements rises correspondingly.
In the end the value of a toolkit such as GoKit is measured less by the number of lines it generates than by the number of decisions it removes from the critical path of feature delivery. Every database, every topic subscription, every continuous-integration step that an engineer no longer has to configure by hand is a unit of attention that can be redirected toward the unique requirements of a partner or a product. When that redirection is systematic and reliable, the organisation as a whole becomes more responsive, more consistent and more capable of sustaining growth without a proportional increase in operational complexity.
The pizza-service narrative also illustrates a secondary benefit that is easy to overlook: the progressive enrichment of operational visibility. Because metrics, dashboards and capability tests are generated rather than hand-crafted, every new service automatically participates in the organisation’s observability and quality regimes. There is no opportunity for a service to be “forgotten” or to ship without the standard instrumentation. Consistency of tooling produces consistency of operational posture, which in turn reduces the cognitive load on both developers and on-call engineers.
Taken together, the practices embodied in GoKit demonstrate that the apparent tension between rapid delivery and long-term maintainability can be resolved by investing in the right abstractions at the platform level. When the platform absorbs the repetitive work of scaffolding, provisioning, testing and observing, individual service teams are free to move quickly without accumulating the structural debt that eventually slows every organisation that scales through pure copy-and-paste. The result is an engineering culture that can sustain high throughput while preserving the coherence and reliability required for a production system that processes hundreds of millions of orders.
The same principles that enabled Jet Connect to move from a maintenance-heavy template model to a coherent, generated estate are available to any organisation facing a similar proliferation of services. The precise technology choices—Terraform versus another infrastructure-as-code system, Kafka versus another event bus, OpenAPI versus another interface description—matter less than the architectural decision to centralise the definition of service shape and to generate the repetitive surrounding machinery. Once that decision is made and enforced, the cost of each additional service falls and the organisation’s ability to respond to new requirements rises. In an environment that processes hundreds of millions of orders, that difference is decisive.
In practical terms the toolkit converts what would otherwise be a multi-day or multi-week effort—creating a repository, wiring continuous integration, declaring infrastructure, adding metrics, writing documentation—into a sequence of minutes. The engineer’s attention remains on the unique aspects of the partner integration or the product feature. Everything else is supplied by generation and enforced by pipeline. That separation of concerns is the essential achievement, and it is the reason the platform can continue to scale without a corresponding explosion in operational overhead.
The experience of Jet Connect demonstrates that the investment in a carefully designed internal platform pays continuing dividends. Each new service inherits the accumulated learning of the entire estate; each improvement to the toolkit propagates automatically; each engineer inherits a consistent, well-instrumented starting point. The result is not merely faster delivery of individual features but a durable increase in the organisation’s capacity to absorb complexity without sacrificing reliability or developer effectiveness.
In short, GoKit is less a code generator than a deliberate architectural intervention that realigns incentives and removes friction from the path of delivery.
Links:
[DevoxxGR2025] Angular Micro-Frontends
Dimitris Kaklamanis, a lead software engineer at CodeHub, delivered an 11-minute talk at Devoxx Greece 2025, exploring how Angular micro-frontends revolutionize scalable web development.
Micro-Frontends Unveiled
Kaklamanis opened with a relatable scenario: a growing front-end monolith turning into a dependency nightmare. Micro-frontends, inspired by microservices, break the UI into smaller, independent pieces, each owned by a team. This enables parallel development, reduces risks, and enhances scalability. He outlined four principles: decentralization (team-owned UI parts), technology agnosticism (mixing frameworks like Angular, React, or Vue), resilience (isolated bugs don’t crash the app), and scalability (independent team scaling). A diagram showed teams building features in different frameworks, integrated at runtime via a shell app.
Pros and Cons
Micro-frontends offer scalability, tech flexibility, faster parallel development, resilience, and easier maintenance due to focused codebases. However, challenges include increased complexity (more coordination), performance overhead (multiple apps loading), communication issues (state sharing), and CI/CD complexity (separate pipelines). Kaklamanis highlighted Angular’s strengths: its component-based structure aligns with modularity, CLI tools manage multiple projects, and features like lazy loading and Webpack 5 module federation simplify implementation. Tools like NX streamline monorepo management, making Angular a robust choice.
Implementation in Action
Kaklamanis demonstrated a live Angular store app with independent modules (orders, products, inventory). A change in the product component didn’t affect others, showcasing isolation. He recommended clear module ownership, careful intermodule communication, performance monitoring, and minimal shared libraries. For large, multi-team projects, he urged prototyping micro-frontends, starting small and iterating for scalability.
Links
[SpringIO2024] Continuations: The Magic Behind Virtual Threads in Java by Balkrishna Rawool @ Spring I/O 2024
At Spring I/O 2024 in Barcelona, Balkrishna Rawool, a software engineer at ING Bank, captivated attendees with an in-depth exploration of continuations, the underlying mechanism powering Java’s virtual threads. Introduced as a final feature in Java 21 under Project Loom, virtual threads promise unprecedented scalability for Java applications. Balkrishna’s session demystified how continuations enable this scalability by allowing programs to pause and resume execution, offering a deep dive into their mechanics and practical applications.
Understanding Virtual Threads
Virtual threads, a cornerstone of Project Loom, are lightweight user threads designed to enhance scalability in Java applications. Unlike platform threads, which map directly to operating system threads and are resource-intensive, virtual threads require minimal memory, enabling developers to create millions without significant overhead. Balkrishna illustrated this by comparing platform threads, often pooled due to their cost, to virtual threads, which are created and discarded as needed, avoiding pooling anti-patterns. He emphasized that virtual threads rely on platform threads—termed carrier threads—for execution, with a scheduler mounting and unmounting them dynamically. This mechanism ensures efficient CPU utilization, particularly in I/O-bound applications where threads spend considerable time waiting, thus boosting scalability.
The Power of Continuations
Continuations, the core focus of Balkrishna’s talk, are objects that represent a program’s current state or the “rest” of its computation. They allow developers to pause a program’s execution and resume it later, a capability critical to virtual threads’ efficiency. Using Java’s Continuation API, Balkrishna demonstrated how continuations pause execution via the yield method, transferring control back to the caller, and resume via the run method. He showcased this with a simple example where a continuation printed values, paused at specific points, and resumed, highlighting the manipulation of the call stack to achieve this control transfer. Although the Continuation API is not intended for direct application use, understanding it provides insight into virtual threads’ behavior and scalability.
Building a Generator with Continuations
To illustrate continuations’ versatility, Balkrishna implemented a generator—a data structure yielding values lazily—using only the Continuation API, eschewing Java’s streams or iterators. Generators are ideal for resource-intensive computations, producing values only when needed. In his demo, Balkrishna created a generator yielding strings (“a,” “b,” “c”) by defining a Source object to handle value yields and pauses via continuations. The generator paused after each yield, allowing consumers to iterate over values in a loop, demonstrating how continuations enable flexible control flow beyond virtual threads, applicable to constructs like coroutines or exception handling.
Crafting a Simple Virtual Thread
In the session’s climax, Balkrishna guided attendees through implementing a simplified virtual thread class using continuations. The custom virtual thread paused execution during blocking operations, freeing platform threads, and supported a many-to-many relationship with carrier threads. He introduced a scheduler to manage virtual threads on a fixed pool of platform threads, using a queue for first-in-first-out scheduling. A demo with thousands of virtual threads, each simulating blocking calls, outperformed an equivalent platform-thread implementation, underscoring virtual threads’ scalability. By leveraging scoped values and timers, Balkrishna ensured accurate thread identification and resumption, providing a clear, hands-on understanding of virtual threads’ mechanics.
Links:
[ScalaDaysNewYork2016] Perfect Scalability: Architecting Limitless Systems
Michael Nash, co-author of Applied Akka Patterns, delivered an insightful exploration of scalability at Scala Days New York 2016, distinguishing it from performance and outlining strategies to achieve near-linear scalability using the Lightbend ecosystem. Michael’s presentation delved into architectural principles, real-world patterns, and tools that enable systems to handle increasing loads without failure.
Scalability vs. Performance
Michael Nash clarified that scalability is the ability to handle greater loads without breaking, distinct from performance, which focuses on processing the same load faster. Using a simple graph, Michael illustrated how performance improvements shift response times downward, while scalability extends the system’s capacity to handle more requests. He cautioned that poorly designed systems hit scalability limits, leading to errors or degraded performance, emphasizing the need for architectures that avoid these bottlenecks.
Avoiding Scalability Pitfalls
Michael identified key enemies of scalability, such as shared databases, synchronous communication, and sequential IDs. He advocated for denormalized, isolated data stores per microservice, using event sourcing and CQRS to decouple systems. For instance, an inventory service can update based on events from a customer service without direct database access, enhancing scalability. Michael also warned against overusing Akka cluster sharding, which introduces overhead, recommending it only when consistency is critical.
Leveraging the Lightbend Ecosystem
The Lightbend ecosystem, including Scala, Akka, and Spark, provides robust tools for scalability, Michael explained. Akka’s actor model supports asynchronous messaging, ideal for distributed systems, while Spark handles large-scale data processing. Tools like Docker, Mesos, and Lightbend’s ConductR streamline deployment and orchestration, enabling rolling upgrades without downtime. Michael emphasized integrating these tools with continuous delivery and deep monitoring to maintain system health under high loads.
Real-World Applications and DevOps
Michael shared case studies from IoT wearables to high-finance systems, highlighting common patterns like event-driven architectures and microservices. He stressed the importance of DevOps in scalable systems, advocating for automated deployment pipelines and monitoring to detect issues early. By embracing failure as inevitable and designing for resilience, systems can scale across data centers, as seen in continent-spanning applications. Michael’s practical advice included starting deployment planning early to avoid scalability bottlenecks.
Links:
[DevoxxFR2015] Scaling Seamlessly with Infinispan on Google Cloud
Ludovic Champenois and Mandy Waite, stepping in for Ray Tsang, delivered a dynamic session at Devoxx France 2015 on Infinispan, a scalable Java-based key/value data store. As Google Cloud Platform advocates, they demonstrated automatic scaling on GCP, showcasing Infinispan’s ability to handle up to 500 nodes effortlessly.
Infinispan’s Scalability Features
Ludovic introduced Infinispan as a highly available data grid, ideal for distributed systems. He explained its key/value store mechanics, optimized for scalability, and demonstrated deployment on GCP’s Compute Engine. The platform’s auto-scaling capabilities adjust resources dynamically, ensuring performance under load.
This flexibility, Ludovic highlighted, simplifies infrastructure management.
Automatic Provisioning and Decommissioning
Mandy detailed GCP’s managed infrastructure, focusing on auto-scaling policies that prioritize removing short-lived or initializing VMs. Q&A clarified mechanisms for controlling instance removal, such as manual group adjustments. This ensures minimal disruption during scaling events, maintaining system stability.
These policies, Mandy noted, enhance operational reliability.
Practical Deployment and Feedback
The duo showcased deploying Infinispan clusters, leveraging GCP’s free trial ($300 credit) for experimentation. They directed attendees to a feedback form and GitHub resources for deeper exploration, encouraging hands-on testing.
This session equips developers for scalable deployments.
Links:
[DevoxxFR2014] Cassandra: Entering a New Era in Distributed Databases
Lecturer
Jonathan Ellis is the project chair of Apache Cassandra and co-founder of DataStax (formerly Riptano), a company providing professional support for Cassandra. With over five years of experience working on Cassandra, starting from its origins at Facebook, Jonathan has been instrumental in evolving it from a specialized system into a general-purpose distributed database. His expertise lies in high-performance, scalable data systems, and he frequently speaks on topics related to NoSQL databases and big data technologies.
Abstract
This article explores the evolution and key features of Apache Cassandra as presented in a comprehensive overview of its design, applications, and recent advancements. It delves into Cassandra’s architecture for handling time-series data, multi-data center deployments, and distributed counters, while highlighting its integration with Hadoop and the introduction of lightweight transactions and CQL. The analysis underscores Cassandra’s strengths in performance, availability, and scalability, providing insights into its practical implications for modern applications and future developments.
Introduction to Apache Cassandra
Apache Cassandra, initially developed at Facebook in 2008, has rapidly evolved into a versatile distributed database system. Originally designed to handle the inbox messaging needs of a social media platform, Cassandra has transcended its origins to become a general-purpose solution applicable across various industries. This transformation is evident in its adoption by companies like eBay, Adobe, and Constant Contact, where it manages high-velocity data with demands for performance, availability, and scalability.
The core appeal of Cassandra lies in its ability to manage vast amounts of data across multiple nodes without a single point of failure. Unlike traditional relational databases, Cassandra employs a peer-to-peer architecture, ensuring that every node in the cluster is identical and capable of handling read and write operations. This design philosophy stems from the need to support applications that require constant uptime and the ability to scale horizontally by adding more commodity hardware.
In practical terms, Cassandra excels in scenarios involving time-series data, which includes sequences of data points indexed in time order. Examples range from Internet of Things (IoT) sensor readings to user activity logs in applications and financial transaction records. These data types benefit from Cassandra’s efficient storage and retrieval mechanisms, which prioritize chronological ordering and rapid ingestion rates.
Architectural Design and Data Distribution
At the heart of Cassandra’s architecture is its data distribution model, which uses consistent hashing to partition data across nodes. Each row in Cassandra is identified by a primary key, which is hashed using the Murmur3 algorithm to produce a 128-bit token. This token determines the node’s responsibility for storing the data, mapping keys to a virtual ring where nodes are assigned token ranges.
To enhance fault tolerance, Cassandra supports replication across multiple nodes. In a simple setup, replicas are placed by walking the ring clockwise, but production environments often employ rack-aware strategies to avoid placing multiple replicas on the same rack, mitigating risks from power or network failures. The introduction of virtual nodes (vnodes) in later versions allows each physical node to manage multiple token ranges, typically 256 per node, which balances load more evenly and simplifies cluster management.
Adding nodes to a cluster, known as bootstrapping, involves the new node randomly selecting tokens from existing nodes, followed by data streaming to transfer relevant partitions. This process occurs without service interruption, as existing nodes continue serving requests. Such mechanisms ensure linear scalability, where doubling the number of nodes roughly doubles the cluster’s capacity.
For multi-data center deployments, Cassandra optimizes cross-data center communication by sending updates to a single replica in the remote center, which then locally replicates the data. This approach minimizes bandwidth usage across expensive wide-area networks, making it suitable for hybrid environments combining on-premises data centers with cloud providers like AWS or Google Cloud.
Handling Distributed Counters and Integration with Analytics
One of Cassandra’s innovative features is its support for distributed counters, addressing the challenge of maintaining accurate counts in a replicated system. Traditional increment operations can lead to lost updates if concurrent clients overwrite each other’s changes. Cassandra resolves this by partitioning the counter value across replicas, where each replica maintains its own sub-counter. The total value is computed by summing these partitions during reads.
This design ensures eventual consistency while allowing high-throughput updates. For instance, if a counter starts at 3 and two replicas each increment by 2, the partitions update independently, and gossip protocols propagate the changes, resulting in a final value of 7 across all replicas.
Cassandra’s integration with Hadoop further extends its utility for analytical workloads. Beyond simple input formats for MapReduce jobs, Cassandra can partition a cluster into segments for operational workloads and others for analytics, automatically handling replication between them. This setup is ideal for recommendation systems, such as suggesting related products based on purchase history, where Hadoop computes correlations and replicates results back to the operational nodes.
Advancements in Transactions and Query Language
Prior to version 2.0, Cassandra lacked traditional transactions, relying on external lock managers like ZooKeeper for atomic operations. This approach introduced complexities, such as handling client failures during lock acquisition. To address this, Cassandra introduced lightweight transactions in version 2.0, enabling conditional inserts and updates using the Paxos consensus algorithm.
Paxos ensures fault-tolerant agreement among replicas, requiring four round trips per transaction, which increases latency. Thus, lightweight transactions are recommended sparingly, only when atomicity is critical, such as ensuring unique user account creation. The syntax integrates seamlessly with Cassandra Query Language (CQL), resembling SQL but omitting joins to maintain single-node query efficiency.
CQL, introduced in version 2.0, enhances developer productivity by providing a familiar interface for schema definition and querying. It supports collections (sets, lists, maps) for denormalization, avoiding the need for joins. Version 2.1 adds user-defined types and collection indexing, allowing nested structures and queries like selecting songs containing the tag “blues.”
Implications for Application Development
Cassandra’s design choices have profound implications for building resilient applications. Its emphasis on availability and partition tolerance aligns with the CAP theorem, prioritizing these over strict consistency in distributed settings. This makes it suitable for global applications where downtime is unacceptable.
For developers, features like triggers and virtual nodes reduce operational overhead, while CQL lowers the learning curve compared to thrift-based APIs. However, challenges remain, such as managing eventual consistency and avoiding overuse of transactions to preserve performance.
In production, companies like eBay leverage Cassandra for time-series data and multi-data center setups, citing its efficiency in bandwidth-constrained environments. Adobe uses it for audience management in the cloud, processing vast datasets with high availability.
Future Directions and Conclusion
Looking ahead, Cassandra continues to evolve, with version 2.1 introducing enhancements like new keywords for collection queries and improved indexing. The beta releases indicate stability, paving the way for broader adoption.
In conclusion, Cassandra represents a paradigm shift in database technology, offering scalable, high-performance solutions for modern data challenges. Its architecture, from consistent hashing to lightweight transactions, provides a robust foundation for applications demanding reliability across distributed environments. As organizations increasingly handle big data, Cassandra’s blend of simplicity and power positions it as a cornerstone for future innovations.
Links:
[DevoxxFR2013] Distributed DDD, CQRS, and Event Sourcing – Part 1/3: Time as a Business Core
Lecturer
Jérémie Chassaing is an architect at Siriona, focusing on scalable systems for hotel channel management. Author of thinkbeforecoding.com, a blog on Domain-Driven Design, CQRS, and Event Sourcing, he founded Hypnotizer (1999) for interactive video and BBCG (2004) for P2P photo sharing. His work emphasizes time-centric modeling in complex domains.
Abstract
Jérémie Chassaing posits time as central to business logic, advocating Event Sourcing to capture temporal dynamics in Domain-Driven Design. He integrates Distributed DDD, CQRS, and Event Sourcing to tackle scalability, concurrency, and complexity. Through examples like order management, Chassaing analyzes event streams over relational models, demonstrating eventual consistency and projection patterns. The first part establishes foundational shifts from CRUD to event-driven architectures, setting the stage for distributed implementations.
Time’s Primacy in Business Domains
Chassaing asserts time underpins business: reacting to events, analyzing history, forecasting futures. Traditional CRUD ignores temporality, leading to lost context. Event Sourcing records immutable facts—e.g., OrderPlaced, ItemAdded—enabling full reconstruction.
This contrasts relational databases’ mutable state, where updates erase history. Events form audit logs, facilitating debugging and compliance.
Domain-Driven Design Foundations: Aggregates and Bounded Contexts
DDD models domains via aggregates—consistent units like Order with line items. Bounded contexts delimit scopes, preventing model pollution.
Distributed DDD extends this to microservices, each owning a context. CQRS separates commands (writes) from queries (reads), enabling independent scaling.
CQRS Mechanics: Commands, Events, and Projections
Commands mutate state, emitting events. Handlers project events to read models:
case class OrderPlaced(orderId: UUID, customer: String)
case class ItemAdded(orderId: UUID, item: String, qty: Int)
// Command handler
def handle(command: AddItem): Unit = {
// Validate
emit(ItemAdded(command.orderId, command.item, command.qty))
}
// Projection
def project(event: ItemAdded): Unit = {
updateReadModel(event)
}
Projections denormalize for query efficiency, accepting eventual consistency.
Event Sourcing Advantages: Auditability and Scalability
Events form immutable logs, replayable for state recovery or new projections. This decouples reads/writes, allowing specialized stores—SQL for reporting, NoSQL for search.
Chassaing addresses concurrency via optimistic locking on aggregate versions. Distributed events use pub/sub (Kafka) for loose coupling.
Challenges and Patterns: Idempotency and Saga Management
Duplicates require idempotent handlers—e.g., check event IDs. Sagas coordinate cross-aggregate workflows, reacting to events and issuing commands.
Chassaing warns of “lasagna architectures”—layered complexity—and advocates event-driven simplicity over tiered monoliths.
Implications for Resilient Systems: Embracing Eventual Consistency
Event Sourcing yields antifragile designs: failures replay from logs. Distributed CQRS scales horizontally, handling “winter is coming” loads.
Chassaing urges rethinking time in models, shifting from mutable entities to immutable facts.
Links:
[DevoxxFR2013] NIO, Not So Simple?
Lecturer
Emmanuel Lecharny is a member of the Apache Software Foundation, contributing to projects like Apache Directory Server and Apache MINA. He also mentors incubating projects such as Deft and Syncope. As founder of his own company, he collaborates on OpenLDAP development through partnerships.
Abstract
Emmanuel Lecharny’s presentation delves into the intricacies of network input/output (NIO) in Java, contrasting it with blocking I/O (BIO) and asynchronous I/O (AIO). Through detailed explanations and code examples, he explores concurrency management, scalability, encoding/decoding, and performance in building efficient servers using Apache MINA. The talk emphasizes practical challenges and solutions, advocating framework use to simplify complex implementations while highlighting system-level considerations like buffers and selectors.
Fundamentals of I/O Models: BIO, NIO, and AIO Compared
Lecharny begins by outlining the three primary I/O paradigms in Java: blocking I/O (BIO), non-blocking I/O (NIO), and asynchronous I/O (AIO). BIO, the traditional model, assigns a thread per connection, blocking until data arrives. This simplicity suits low-connection scenarios but falters under high load, as threads consume resources—up to 1MB stack each—leading to context switching overhead.
NIO introduces selectors and channels, enabling a single thread to monitor multiple connections via events like OP_READ or OP_WRITE. This non-blocking approach scales better, handling thousands of connections without proportional threads. However, it requires manual state management, as partial reads/writes necessitate buffering.
AIO, added in Java 7, builds on NIO with callbacks or futures for completion notifications, reducing polling needs. Yet, it demands careful handler design to avoid blocking the callback thread, often necessitating additional threading for processing.
These models address concurrency differently: BIO is straightforward but resource-intensive; NIO offers efficiency through event-driven multiplexing; AIO provides true asynchrony but with added complexity in callback handling.
Building Scalable Servers with Apache MINA: Core Components and Configuration
Apache MINA simplifies NIO/AIO development by abstracting low-level details. Lecharny demonstrates a basic UDP server: instantiate IoAcceptor, bind to a port, and set a handler for messages. The framework manages buffers, threading, and protocol encoding/decoding.
Key components include IoService (for acceptors/connectors), IoHandler (for events like messageReceived), and filters (e.g., logging, protocol codecs). Configuration involves thread pools: one for I/O (typically one thread suffices due to selectors), another for application logic to prevent blocking.
Scalability hinges on proper setup: use direct buffers for large data to avoid JVM heap copies, but heap buffers for small payloads in Java 7 for speed. MINA’s executor filter offloads heavy computations, maintaining responsiveness.
Code example:
DatagramAcceptor acceptor = new NioDatagramAcceptor();
acceptor.setHandler(new MyHandler());
SocketAddress address = new InetSocketAddress(port);
acceptor.bind(address);
This binds a UDP acceptor, ready for incoming datagrams.
Handling Data: Encoding, Decoding, and Buffer Management
Encoding/decoding is pivotal; MINA’s ProtocolCodecFilter uses encoders/decoders for byte-to-object conversion. Lecharny explains cumulative decoding for fragmented messages: maintain a buffer, append incoming data, and decode when complete (e.g., via length prefixes).
Buffers in NIO are crucial: ByteBuffer for data storage, with position, limit, and capacity. Direct buffers (allocateDirect) bypass JVM heap for zero-copy I/O, ideal for large transfers, but allocation is costlier. Heap buffers (allocate) are faster for small sizes.
Performance tests show Java 7 heap buffers outperforming direct ones up to 64KB; beyond, direct excels. UDP limits (64KB max) favor heap buffers.
Partial writes require looping until completion, tracking written bytes. MINA abstracts this, but understanding underlies effective use.
public class LengthPrefixedDecoder extends CumulativeProtocolDecoder {
protected boolean doDecode(IoSession session, IoBuffer in, ProtocolDecoderOutput out) {
if (in.remaining() < 4) return false;
int length = in.getInt();
if (in.remaining() < length) return false;
// Decode data
return true;
}
}
This decoder checks for complete messages via prefixed length.
Concurrency and Performance Optimization in High-Load Scenarios
Concurrency management involves separating I/O from processing: MINA’s single I/O thread uses selectors for event polling, dispatching to worker pools. Avoid blocking in handlers; use executors for database queries or computations.
Scalability tests: on a quad-core machine, MINA handles 10,000+ connections efficiently. UDP benchmarks show Java 7 20-30% faster than Java 6, nearing native speeds. TCP may lag BIO slightly due to overhead, but NIO/AIO shine in connection volume.
Common pitfalls: over-allocating threads (match to cores), ignoring backpressure (queue overloads), and poor buffer sizing. Monitor via JMX: MINA exposes metrics for queued events, throughput.
Lecharny stresses: network rarely bottlenecks; focus on application I/O (databases, disks). 10Gbps networks outpace SSDs, so optimize backend.
Practical Examples: From Simple Servers to Real-World Applications
Lecharny presents realistic servers: a basic echo server with MINA requires minimal code—set acceptor, handler, bind. For protocols like LDAP, integrate codecs for ASN.1 encoding.
In Directory Server, NIO enables handling massive concurrent searches without thread explosion. MINA’s modularity allows stacking filters: SSL for security, compression for efficiency.
For UDP-based services (e.g., DNS), similar setup but with DatagramAcceptor. Handle datagram fragmentation manually if exceeding MTU.
AIO variant: Use AsyncIoAcceptor with CompletionHandlers for callbacks, reducing selector polling.
These examples illustrate MINA’s brevity: functional servers in under 50 lines, versus hundreds in raw NIO.
Implications and Recommendations for NIO Adoption
NIO/AIO demand understanding OS-level mechanics: epoll (Linux) vs. kqueue (BSD) for selectors, impacting portability. Java abstracts this, but edge cases (e.g., IPv6) require vigilance.
Performance gains are situational: BIO suffices for <1000 connections; NIO for scalability. Frameworks like MINA or Netty mitigate complexity, encapsulating best practices.
Lecharny concludes: embrace frameworks to avoid reinventing; comprehend fundamentals for troubleshooting. Java 7+ enhancements make NIO more viable, but test rigorously under load.
Relevant Links and Hashtags
Links:
[DevoxxFR2013] Web Oriented Architecture: A Transmutation of Practices in Building Information Systems
Lecturer
Habib Guergachi is a Centrale graduate and the CEO and co-founder of Zenexity, established in 2003. As an expert in urbanization, he is among the French architects who introduced key concepts such as the “integrability coefficient” of applications. His professional background includes over seven years at the Central IT Department of AXA, more than three years at the IT Strategy Department of Société Générale, and five years on the executive committee and Technical Direction of a major IT services company. Currently, he leads large-scale urbanization projects and transformations of information systems toward web-oriented models. He also conducts seminars at the prestigious Capgemini Institute.
Abstract
This article explores Habib Guergachi’s lecture on Web Oriented Architecture (WOA), which critiques traditional enterprise practices and advocates for a shift to distributed, hyper-scalable systems. Drawing from historical analogies and real-world examples like LinkedIn, Guergachi argues for abandoning monolithic architectures in favor of independent, reactive applications that leverage modern web protocols. The discussion analyzes the implications for software development, emphasizing innovation, scalability, and the rejection of outdated paradigms to ensure future competitiveness in the French and global IT landscape.
Contextualizing the Shift from Hyper-Integrated to Hyper-Distributed Systems
Guergachi begins by drawing a parallel between the decline of traditional industries, such as steel mills like Gandrange and Florange, and the potential obsolescence of current software engineering practices. He posits that modern IT specialists, akin to specialized workers in software factories, risk irrelevance if they fail to innovate. The core dilemma is the overemphasis on hyper-integrated systems, where enterprises purchase off-the-shelf software that imposes architectures dictated by vendors. This leads to rigid, costly structures that stifle adaptability.
In contrast, Guergachi introduces the concept of hyper-distributed architectures, inspired by web-oriented principles. He illustrates this with a cultural anecdote: a hypothetical Chinese viewer searching for a French film clip on ina.fr fails due to rigid, integrated search mechanisms, while Google succeeds through flexible, distributed intelligence. This highlights how integrated systems, often built around enterprise architecture, application servers, and service buses, create “bousins” – complex, unmaintainable assemblages of tools like CMS for content, transactional plugins, and adapters for legacy JSF applications.
The lecturer critiques the inefficiency of such systems, where decision-making processes involve dumping data into warehouses for analysis, rather than fostering real-time adaptability. He urges a generational shift: respecting past achievements that built foundational information systems but making way for younger developers to construct future-proof ones. Avoiding the trap of using ingenuity merely to bypass imposed integrations is crucial, as technological evolution accelerates.
Principles of Distributed Paradigms and Real-World Implementations
Central to Guergachi’s thesis is the advocacy for distributed paradigms over integrated ones. He references Play Framework, a French-origin technology (despite initial skepticism due to its nationality), as a tool for building independent applications. LinkedIn’s approach exemplifies this: constructing systems as separate components, each focused on core business logic, deployed autonomously. These might use various technologies but prioritize scalability, security, and reactivity.
In a distributed model, non-core functions like search are outsourced to specialized services, allowing internal applications to remain focused and resilient under load. Guergachi explains techniques such as eventual consistency for high-load scenarios and strict consistency only where necessary, like payment processing. Communication between applications relies on RESTful hypermedia over HTTP, rejecting heavy RPC protocols or plugins like Flash, which he derides as symptomatic of a “third-world syndrome” – adopting external technologies without deep understanding.
He envisions enterprises concentrating solely on core business, externalizing storage, CMS, back-office, and video management to superior providers. Performance concerns with HTTP are dismissed as psychological barriers; no in-house solution can compete with storage specialists. Applications will interconnect in a “spaghetti” manner, but one that ensures predictability and adaptability, mirroring the web’s organic structure.
Guergachi introduces entropy as a metaphor: solid (rigid, controlled architectures), liquid (flexible, scalable across servers), and gaseous (pervasive, occupying value chain interstices like Google). Enterprises must evolve toward gaseous states for survival, contrasting with legacy systems that “suck blood” through perpetual maintenance fees.
Implications for Innovation and the Role of French IT Genius
The lecturer delineates integrated paradigms – building overarching technical architectures without functional hypotheses, aiming for longevity – as flawed, akin to overpacking for unforeseen disasters. Distributed paradigms, conversely, tailor architectures per application, prioritizing functional solutions over technical absolutes. For instance, displaying cached content during network failures ensures usability, decided by business logic rather than rigid transactional rules.
A paradigm, per Guergachi, is a coherent worldview offering solutions to its own problems. He warns against half-measures, like deploying advanced frameworks on outdated servers, urging full commitment to distributed models despite risks like dismissal. Submitting to vendor-driven technologies prepares for shameful obsolescence, whereas bold shifts enable glory through innovation.
Critiquing entities like INPI’s outdated systems, he highlights France’s image issues, comparable to 1980s Korean cars. French IT genius, exemplified by talents like Guillaume Bort and Sadek Drobi, must harness business acumen. Concepts like Scalaz originated in France (at Camel), underscoring untapped potential.
The economy of the web remains to be fully realized; Silicon Valley leads but hasn’t won. French informatics must act through innovation serving functionality, user experience, and distributed architectures with increasing entropy. Mastering interconnection complexity yields value, constructing planetary software masterpieces to safeguard jobs and elevate France.
In conclusion, Guergachi’s call to action – rebooting mindsets Monday morning – emphasizes radical change for continuity. By embracing WOA, developers transcend traditional constraints, fostering systems that are open, secure, adaptable, and cost-effective, aligning with business evolutions.